In a traditional content delivery/distribution network, the contents are stored in one or more centralized locations, and then distributed to the endpoints from centralized servers. The contents are distributed either directly or through a pre-arranged proxy server. This arrangement works well in a small-scale system where the contents are mostly static.
However, when the contents are updated very often, and when the set of users and the distribution network are very large, both the distribution of the contents among the servers and medium and long-haul access to these servers for up-to-date contents cause several problems. For example, this causes an excessive amount of traffic in the transport network which leads to excessive delay in downloading the contents because of congestion in the transport segment of the network.
In order to overcome these problems, researchers have experimented with the concept of distributed hash tables (DHT) for peer-to-peer communications and file sharing. DHT relieves the system of the need for centralized servers by using an algorithmic approach for efficient routing to disseminate control information related to the content or file from peer-to-peer. The DHT approach, therefore, helps to achieve an improved level of scalability and storage.
However, because DHT does not use keyword searches to find the desired information or files, the probability of successfully finding the exact desired information varies significantly. The probability of success can be improved if specific information were to be employed. However, the specific information required is rarely available. Because of the use of different virus protection mechanisms in each peer, the downloaded files could inadvertently bring in serious security threats to the hosts and the mobile devices.